Childhood brain tumors (CBT) are the most common solid tumor in children ages 0-19. Epidemiologic studies of CBT done to date suggest that long suspected BT risk factors such as ionizing radiation, head trauma and known genetic predispositions account for only a small proportion of incident cases. A compelling experimental model suggests that N-nitroso compounds (NOC) may cause BT in humans, particularly when exposure is transplacental. Since the tumors that are induced by transplacental NOC exposure of non-human primates occur in young adult as well as in immature animals it seems reasonable in our CBT study to include all cases diagnosed to age 20. This proposed U.S. study, in conjunction with the concurrent international study, will be the first to focus on the NOC/CBT hypothesis. We propose to study all cases of primary BT diagnosed from 1984- 1990 among children ages 0-19 living in all areas of the West Coast that are covered by population-based tumor registries; this includes 19 counties in the Los Angeles (LA), San Francisco (SF) and Seattle areas. In both SF and Seattle we expect to interview the biological mothers of 105 children with BT diagnosed during this 7 year period; in LA we expect to interview 300 case mothers. In each region random digit dial controls stratified by age and sex will be selected with a control:case ratio of 2:1 for SF and Seattle and 1:1 in LA. We expect to complete interviews with mothers of 510 CBT cases and 720 controls and with 1000 fathers. The questionnaire asks about sources of NOC exposure such as diet, drinking water, drugs, cosmetics, rubber products and parental occupational exposures. Pathology slides of all cases will be reviewed by a single pediatric neuropathologist. Data for each of the 3 West Coast regions will be analyzed separately and will also be pooled. The combined West Coast study will have sufficient statistical power to conduct separate analyses for the 2 most common histologic types of CBT- astrocytoma and medulloblastoma. Data will also be comparable with data from a concurrent international study. Data from the proposed West Coast study will be combined with data from the international study to enable separate analyses for several of the less common histologic types.